INDEX

Explanations

terms related to directionality, specifically the words "upstream" and "downstream."

oai_token-act-pair · gpt-3.5-turbo

context related to upstream and downstream processes or events

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iasis

-0.81

ces

-0.74

thritis

-0.73

itia

-0.72

neum

-0.68

 Manson

-0.68

ter

-0.67

inea

-0.67

nov

-0.67

html

-0.66

POSITIVE LOGITS

 downstream

1.49

 upstream

1.47

é¾

0.78

stream

0.76

 cascade

0.76

 prelim

0.75

eatures

0.74

 mosqu

0.73

river

0.73

edIn

0.72

Activations Density 0.009%

terms related to directionality, specifically the words "upstream" and "downstream."

context related to upstream and downstream processes or events

No Comments

No Known Activations